Joint user pairing, resource block allocation, and power control for full-duplex cellular networks

Abstract

Full-duplex (FD) communication can be applied to the cellular network to improve the spectral efficiency, which has been recently proved feasible because of the advanced self-interference cancellation techniques. In FD communication enabled cellular networks, how to effectively pair a downlink user and an uplink user to transmit on the same wireless channel is a critical and challenging issue. To address it, we formulate a non-convex combinatorial optimization problem to maximize the overall system throughput by user paring, resource block allocation, and power control in a long-term evolution (LTE) system. Due to the NP-hardness of the problem, we first relax the integer variables into continuous ones and then utilize the concave-convex procedure to develop an upper-bound algorithm. To reduce the computational complexity, we also propose a heuristic algorithm which decomposes the problem into three sub-problems: user pairing, resource block allocation, and power control. Numerical results show that FD is superior to HD when the residual self-interference is small enough and the heuristic algorithm can achieve a close performance as the upper-bound algorithm.